Two researchers from Singapore are retracting a paper that included errors in three figures because there’s apparently no way to fix the mistakes and have the new work reviewed.
Here’s the notice for “Host-dependent effects of the 3′ untranslated region of turnip crinkle virus RNA on accumulation in Hibiscus and Arabidopsis,” by Weimin Li and Sek-Man Wong of National University of Singapore and Temasek Life Sciences Laboratory:
The authors regret to inform the readers of Journal of General Virology that honest errors were made in the preparation of Fig. 1(c), Fig. 2 and Fig. 3(b) (the rRNA controls were either missing or placed wrongly), such that the findings as presented are deemed unreliable. While the authors stand by the conclusions of the paper, there is currently no editorial mechanism to review the errors. The authors are currently repeating the experiments and will resubmit the paper to Journal of General Virology in due course, but agree to retract the paper at this time, and they regret any inconvenience caused to the readers of the journal.
The paper has been cited four times, according to Thomson Scientific’s Web of Knowledge.
We’re not quite sure what “no editorial mechanism to review the errors” means, nor why these errors came to light more than five years after the paper was published, but we’ve asked the editor of the journal and will update with anything we learn.
Update, 12 p.m. Eastern, 3/13/13: Editor Stacey Efstathiou, who took over as editor of the journal last year, sent us this detailed response (It turns out that our streams had crossed in reporting this post, and Efstathiou had sent us this last month when we first asked. Apologies for that mixup):
The Editor in Chief of JGV, Professor Richard Elliott (EiC 2007-2012) was contacted on 22nd October 2012 by a scientist working in the field who identified anomalies with certain gel images published in the above manuscript. Dialogue with the corresponding author of the above paper was initiated on 24th October 2012. In conjunction with input from the Editor who had handled the paper in 2007, the Editor in Chief investigated the alleged anomalies and concluded that certain data in this publication were unreliable. Following Committee on Publication Ethics (COPE) guidelines a decision was taken that the paper should be retracted. The EiC contacted the corresponding author on 23rd November 2011 with suggested wording for a retraction. The authors wished to repeat relevant experiments with appropriate controls in order to verify their published data and publish a correction or, if necessary, a retraction. On 10th December 2012 the Journal insisted on a retraction.The sentence in the retraction statement referred to in your email reflects the fact that JGV follows COPE guidelines, and has no additional process in place to review data where authors request to repeat their experiments and publish a correction to their original paper.JGV strongly supports COPE, which recommend retraction in cases where published findings are unreliable. We feel that the COPE guidelines are transparent and fair, hence the approach taken in this case. Retraction and, where appropriate, resubmission is in our view a more rigorous approach.
I understand something like the editorial board didn’t come to a consensus on wheter they shoudl accept a new figure/correction or retraction then ressubmission. The second option seems ridiculous if the data will be pretty much the same but with some new figures (maybe just one!).
Well if the editors don’t have a mechanism to figure it out, maybe they need some help? Here’s the original: http://vir.sgmjournals.org/content/94/Pt_3/694.full
In Figure 1c, the left lane with all its beautiful distinguishing features, is a copy of the first lane from Figure 2c, flipped horizontally and re-sized vertically. It’s not the RNA controls as the authors contend, but the entire lane of data. The loading controls beneath the panels are different in each case. These are at least the same experimental conditions, so if we assume the horizontal flip was “accidental” then this might be excused as a genuine mistake.
In Figure 3b, the left lane of the left panel is copied from one of the lanes in Figure 3c (right panel, right lane). The loading controls beneath these panels are different in each case. The experimental conditions represented by these 3 panels are completely different. The image has intensity (saturation) has been adjusted slightly between the panels. Maybe this was a mistake too?
The authors are going to repeat an experiment more than five years after it was done and published? Does not that, somehow, sound difficult? (without a generous funding source, that is)